Video Coding Skip State Based on Subunit Difference

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current High Efficient Video Coding (HEVC) methods face high computational complexity and low video encoding efficiency due to inflexible intra-prediction mode selection processes, which increase the difficulty in real-time communication.

Innovation Solution

A video encoding method that determines a skip state for intra-prediction mode based on the degrees of difference between subunits, allowing for the skipping of intra-prediction mode execution and selection of inter-prediction mode when appropriate, thereby reducing computational operations and improving encoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If intra-prediction mode selection is performed for each coding unit, then video encoding quality is improved, but computational complexity increases

Engineering Contradiction:
Improvevideo encoding qualityVSAvoidcomputational complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The current coding unit is divided into multiple subunits, and the degree of difference between adjacent subunits is calculated. This segmentation allows the system to analyze local variations within the coding unit and make more informed decisions about mode selection, improving encoding quality while managing complexity through hierarchical analysis.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent calculates the degree of difference between subunits before performing full intra-prediction mode selection. This preliminary analysis of subunit similarity allows the system to pre-determine whether intra-prediction is likely to be beneficial, avoiding unnecessary computational operations and reducing overall complexity.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If intra-prediction mode is selected for each coding unit, then encoding accuracy is improved, but encoding efficiency decreases

Engineering Contradiction:
Improveencoding accuracyVSAvoidencoding efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent introduces a skip mechanism where, based on the calculated degree of difference between subunits, the system can skip performing intra-prediction mode selection entirely for certain coding units. When the degree of difference exceeds a threshold, the system directly selects inter-prediction mode, skipping the computationally intensive intra-prediction analysis while maintaining acceptable encoding accuracy.

Inventive Principle:
Principle #21Skipping (Rushing through)

Solution Approach 2:

The patent uses the degree of difference between subunits as a dynamic parameter to control the encoding process. By changing this parameter (comparing it against thresholds), the system adaptively adjusts its behavior between performing full intra-prediction analysis and skipping directly to inter-prediction, thereby balancing accuracy and efficiency based on local image characteristics.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If flexible mode selection is implemented, then video encoding efficiency is improved, but computational operations increase

Engineering Contradiction:
Improvevideo encoding efficiencyVSAvoidcomputational operations
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent performs a preliminary calculation of the degree of difference between subunits before committing to full mode selection. This preliminary action provides a quick indicator of whether flexible mode selection is likely to benefit the current coding unit, allowing the system to avoid unnecessary computational operations while maintaining the flexibility to adapt to local image characteristics.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3748966B1Video coding method, computer device, and storage medium
Publication Date: 2024.02.21 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP3748966B1 patent drawingFigure 1~3
  • EP3748966B1 patent drawingFigure 4~5
  • EP3748966B1 patent drawingFigure 6~7

AI summary

This application relates to a video coding method, the method comprising: acquiring a subunit set corresponding to a current coding unit, wherein the subunit set is composed of subunits obtained by pre-dividing the current coding unit; according to the degree of difference between the subunits in the subunit set, determining a skip state of an intra-prediction mode corresponding to the current coding unit; and when the skip state of the intra-prediction mode gives a positive indication, skipping performing the intra-prediction mode on the current coding unit, and coding the current coding unit according to the inter-prediction mode.